THE SPERMATOGENESIS OF DOMESTIC MAMMALS. 395 
the loops approximate and fuse more or less completely. In this condition 
the bivalent chromosome is divided in metakinesis.” 
On the question of synapsis and reduction in Hemiptera Tomopteris, 
Bastracoseps and some other forms Wırson (712) has published an excellent 
and critical consideration in which he stated that: “ The cytological problem 
of synapsis and reduction involves four principal questions as follows: (1) 
Is synapsis a fact? Do the chromatin-elements actually conjugate or other- 
wise become associated two by two? (2) Admitting the fact of synapsis, are 
the conjugating elements chromosomes, and are they individually identical 
with those of the last diploid or pre-meiotie division ? (3) Do they conjugate 
side by side (parasynapsis, parasyndesis), end to end (telosynapsis, metasynde- 
sis) or in both ways? (4) Does synapsis lead to partial or complete fusion 
of the conjugating elements to form ‘ Zygosomes ’ or ‘ Mixochromosomes,’ or are 
they subsequently disjoined by a ‘reduction division’ ? Upon these questions 
depends our answer to a fifth and still more important question, namely, (5) 
Can the Mendelian segregation of unit factors be explained by the phenomena 
of synapsis and reduction ?” 
In Bos taurus, however, it is impossible to consider these whole questions 
of synapsis and reduction. As already described, in the telophase of the 
ultimate spermatogonia the chromosomes do not fuse but remain for a while 
scattered throughout the nucleus. In these cells thirty or more chromosomes 
can be counted. In postsynapsis (pachytene stage) the spiremes appear in 
about half the original number. ‘These conditions clearly indicate that in the 
presynapsis the chromosomes are in the diploid number and conjugate during 
the synaptene stage. 
It is difficult to determine whether the conjugation of chromosomes takes 
place by parasynapsis (side by side conjugation) or telosynapsis (end to end con- 
jugation). In the horse, from many phenomena, we have reached the view that 
the conjugation of chromatin threads probably takes place by parasynapsis. 
But in cattle there is no evidence of parasynapsis taking place at all. During 
the synaptene stage a longitudinal duality of the spiremes, as that seen in 
the horse, is almost absent or, if occasionally present, is due either to an 
accidental parallelism, or to a longitudinal split comparable to that of the 
diploid prophase. But as stated above, the observation of the preparations 
